Just for fun, I printed out Anthromod’s Kickstarter Hand finger:
If I were doing it, I’d add 1.75 mm alignment holes in each part, but clamping each phalange in both directions came out close enough:
The tolerances were a bit tight and it required some trimming before all the joints flexed freely. I used short segments of 3 mm orange filament for the knuckle hinges and heat-staked the ends, rather than having to trim a trio of 3 mm screws:
After making three short rubber bands by tying and trimming loops from a longer band, the finger curled up just like yours:
The overall quality isn’t as good as I’d like: there’s a bit of uplift on the edges and corners. If I print another one, I want to try less than 0.2 infill and less cooling.
The Smell of Molten Projects in the Morning 
Certainly looks like fun. I have held off on the 3D printing movement letting all the bugs get worked out and waiting for the resolution to improve, well that and lack of time and space. I continue to use subtractive processes to make what I need but certainly look forward to using additive at some point for the more complicated and even impossible to machine parts. I do not currently have CNC but rather enjoy the additional skill it takes to use the machines manually. I do actually need/want CNC, just have not gotten there yet. I firmly believe manual machining makes for a better CNC programmer.
And I believe that knowing how to program in assembly makes for a better programmer. But try to tell the kids these days that…
I have had that discussion with a certain Larval Engineer; seems they don’t even mention that stuff these days, because it’s just Not Relevant. Given that they’ve outlawed pointers in all the fancy new languages, knowing how the machinery actually fetches operands doesn’t matter…
So I’m told, anyhow. Sign of the times, sez I.
Hah, I bet it would be hard to convince most of them too. Even though it has been as while, I’m certain I could pick it up again.
Absolutely! I can hammer out G-Code just fine, but my fingers don’t have the smarts that know the proper speeds-and-feeds; that G-Code is no more than adequate. Fortunately, I’m not doing production work and close enough is good enough.
On the other paw, having a 3D printer is wonderful beyond compare. Yesterday I needed a bracket (more on this later) that “looks like this” with holes here-and-there… and it magically appeared on the platform. I could gnaw it out of an aluminum (or polycarb) slab, but it’s ever so much easier to conjure an OpenSCAD model and have it Just Work. Particularly because
ifwhen the measurements are slightly off, I can get another one that’s just like the other one, only slightly different in exactly one dimension, without any additional hassle.I like how the printing layers look a little like fingerprints.
They do, even with that slot through the middle!
Of course, that’s a defect for some folks, but I kinda like the wood-grain effect on most objects…
“On the other paw, having a 3D printer is wonderful beyond compare.”
I know, I know, but I just can’t get past the look of the layers not being well hidden, aka low resolution. I could show someone (who is not into making/repairing/tinkering) a part from a home-built 3D printer and tell them I made it and they might believe it, however if I made one from a block of nylon, aluminum, steel, etc. and gave it a suitable surface finish, then they usually don’t believe me … that’s when I know I’ve done a great job. Certainly they have their place and the resolution will get better or more affordable but for now I not in any rush. But I do get your point of how quickly you can have the part versus what I have to go through to get the same thing. I have had my eye on the UV curable versions … someday. (See: http://www.kickstarter.com/projects/formlabs/form-1-an-affordable-professional-3d-printer)
I use 0.25 mm layers, mostly because it works out evenly for the typical hard-metric dimensions I use, but folks who care about surface finish routinely run 0.10 and under, with zanies under 0.05. Word has it that, at that level, the side striations pretty much Go Away, although there’s certainly contouring on a non-planar top surface. There’s also an obvious inverse relationship between layer thickness and printing time that factors into my tolerance for relatively coarse layers.
The vapor-bath technique of smoothing the surfaces seems to work well with ABS, moderately with PLA, and poorly with anything else. How it affects precise dimensions remains to be seen; I’m reluctant to fiddle with acetone or tetrahydrofuran vapor in the Basement Laboratory.
The DIY-scale laser-sintered and UV-cured “printers” have interesting possibilities that are, I think, severely limited by patent issues; Formlabs seems to be settling a lawsuit by 3D Systems, with few details having emerged.
Re vapor baths: When I was a larval engineer in the ’70s, the Moto research lab I interned at did a simple DIY vapor degreaser. Commercial ones use liquid (or refrigerant) cooling, but we used a simple aluminum heat sink. Take a tall beaker (1 liter, I think, 100mm by 200mm), and length of alumimum flashing 8″ wide. Cut fins in one half and stick the intact portion in the beaker, to get an upside down tophat. (Yes, I do mix measurement systems. It’s part of my heritage.)
Put the beaker on a lab type hot plate, put an inch or two of solvent in, and set heat to get a vapor of the solvent. We covered the beaker with aluminum foil when not in use.
IIRC, we used TCE (still legal to use then) and a proprietary solvent in the vapor degreasers, but acetone should work. You need to set the heat just enough to get a good vapor, but when set right, you don’t get obnoxious fumes. Best to experiment in a garage, assuming no flame sources there. In a lab situation, I’d want some venting with a no-spark fan…
And a loooong stick. Selah.